Mamdani fuzzy rule based model to classify sites for aquaculture development

Abstract

Mamdani fuzzy inference system was applied as a decision making model to classify aqua sites based on water, soil, support, infrastructure, input, and risk factor related information. For input and output linguistic variables of the model, suitable Gaussian and triangular membership functions were selected. Totally, 729 rules with logical AND operator, truncation implication, and centriod method for defuzzification were employed to develop an efficient fuzzy model for decision making about classification of aqua sites. The model classifies each site in the datasets into one of the three classes such as suitable, moderate or unsuitable. In order to validate the performance of the proposed fuzzy model, the same sets were classified again by aquaculture expert. Classification results obtained from the developed fuzzy model showed 92% agreement with the results from the aquaculture expert. Thus the fuzzy rule based model is a feasible model for classification of aqua sites, it involves less computation and has clear implementation and working schemes.